There can be little doubt that royal women had their own apartments (Briant 2002: 283-4) either in tents, wagons, or palaces; Herodotus (3.68-9, 7.2-3) certainly thought as much when he described the physical layout of the inner court of Smerdis and the book of Esther (2:13, 16) also envisages separate living spaces for women. When the Greek doctor Democedes arrived at court he was escorted by a eunuch to meet the king’s wives (Herodotus 3.130) and we also hear that ‘before the age of five a boy lives with the women and never sees his father’ (Herodotus 1.136) - although this is no doubt something of an exaggeration, in line with other Greek texts on the harmfulness of a ‘harem rearing’ (D1).
What can be done with these observations? In his History of the Persian Empire, Olmstead paints an atmospheric picture of an L-shaped building at the southern edge of Persepolis (F19) which Herzfeld (1941) and Schmidt (1953) identified as the harem:
Surrounded by the guardrooms of the watchful eunuchs was a tier of six apartments to house the royal ladies. Each tier consisted of a tiny hall whose roof was upheld by only four columns and a bedroom so minute that even with a single occupant the atmosphere must have been stifling. (Olmstead 1948: 285)
Such a description makes it unsurprising that many scholars have found it difficult to accept this section of the terrace as the space occupied by the Empire’s foremost women, certainly when compared with the Greek and Hebrew texts, although few scholars have analysed the physical remains of the structures in any detail. Brosius (2007: 33) dismisses Herzfeld’s and Schmidt’s designation of the area out of hand: ‘so far no structure has been identified at Persepolis which could have served as the women’s quarters’. Indeed, the building has often been classified as an overflow storeroom of the nearby treasury (Wilber 1969: 73).
The position of the L-shaped building towards the back of the terrace provides strong support for it being a (temporary) residential area of part of the harem. Add to this a number of uniform apartments within the complex, each consisting of a main room connected with one smaller room or two such subsidiary chambers, and it would appear that the argument for a living space is enhanced. These were certainly the principal criteria for the structure’s initial identification as the harem (Schmidt 1953: 255; supported by Shahbazi 2004: 163), although doubts have been cast on Herzfeld’s initial motives in identifying the remains as such. Allen notes that at the time of the first publication of its discovery in the 1930s, Herzfeld was in the process of negotiating funds for the excavations and that the potential recovery of a harem was held to be propitious because of the hope (unfulfilled) of finding precious artefacts within the remains (Allen 2007: 329).
Let us reconsider the location of the harem on the terrace. It lies well inside the area of the platform defined as private. In fact, Herzfeld argued for the strict separation of this area from those accessible to the public, on the model of similar layouts of the majority of other ancient Near Eastern palaces (Herzfeld 1941: 226; Allen 2007: 328). A glance at the plan of Persepolis and other Near Eastern residential palace quarters (at, for instance, Babylon and Nimrud) reveals that these buildings contain individual rooms of modest scale, certainly in comparison with the monumental grandeur of the areas intended for public display. At Persepolis the protection of these structures by the thick southern fortification wall immediately behind the harem contributes to the function proposed here for the building, as one would expect to find accommodation used by the king and the royal family to be best protected; indeed, the presence of guard reliefs at major entrances to the compound suggest that security was paramount (Root 1979: 10) (the figures of soldiers carved into the connecting wall between the upper terrace and the harem are usually overlooked by scholars). Crucially then, this space at the terrace rear, which was allocated as living quarters for at least some of the royal family, was hidden by high fortifications and was well guarded by the military. It was secure and private.
The harem is grouped with other palatial residential structures both on and off the platform and it is actually integral to the building immediately above it - identified as Xerxes’ palace and private residence. Xerxes’ palace is connected to the harem by two grand, well worked flights of stairs, which must have been utilised by the king or his courtiers when they required direct access to the rooms below (F19; Schmidt 1953: 244). They could move between the two parts of the palace without having to traverse any public space. Schmidt’s excavations found that the lower flights of steps were formerly enclosed, while the upper section was open, and that the more monumental, well dressed, and polished western stairway also contained one of the few physically evidenced (well worked) doors in the area. Schmidt also identified a direct access route, via the stairways, connecting the harem with the Council Hall and the Hall of a Hundred Columns, allowing the king and the royal family to move conveniently and directly from their private apartments to the public areas without breaching security (Schmidt 1953: 255).
As to the apartments themselves, they are laid out in two rows, all interconnected by long narrow corridors. From archaeology (F19) one can identify a maximum of twenty-two apartments, each consisting of a large hypostyle room with one or more adjoined chambers (Herzfeld 1941: 229; Schmidt 1953: 137, 260). The main rooms are well decorated with niches and plastered walls and elegant stone door lintels and column bases, which are so well crafted that the general execution of the stonework is just as fine as that on the palaces of the kings (Wilber 1969: 94). The average apartment measures approximately ten by ten metres - and while in no way a negligible living space, it would be hard to imagine royal personages passing their days perpetually in a room this size. This alone helps to negate the image of royal women living in strictly guarded confinement - the claustrophobia would have been cruel - and Briant rightly voices his concern that it is not at all clear ‘that the royal princesses lived cloistered in their apartments’ (Briant 2002: 285). It is better to think of each chamber as perhaps a separate (and temporary) domestic quarter (for sleeping?) or as an antechamber or storage area but not as a room used by a single occupant all of the time.
A big courtyard in the harem’s main wing and the large room attached to it are therefore best interpreted as a communal space for the harem rather than for a grand individual’s private use (as Schmidt proposed), since it lacks the domestic quarter/antechamber units which accompany the main halls of Darius’ and Xerxes’ palaces. In view of the regulated control of movement to this area, both from within and outside the harem, it could have functioned as an audience chamber for royal women or princes or perhaps even the king when he chose to remain there. Interestingly, what appears to be a female audience is depicted on a cylinder seal (F20). The parallel with the motif of the king’s audience (F3) is explicit and is proof of the high regard in which royal women - possibly in this instance the king’s mother - were held (Brosius 1996: 86; Brosius 2010a; Lerner 2010). It is very likely that in this large hall a space was set aside for other activities, including communal eating and entertainment, as well as the collective rearing of younger children; we must certainly be rid of any notion of women shut up all day in cramped, isolated cells. The cluster of defined ‘apartments’ certainly accords with Diodorus’ description of residences on the terrace (7.70-1): ‘Scattered about the royal terrace were apartments of the kings and members of the royal family as well as quarters for the great nobles’. Variations within the standard model of the chambers may provide a convenient indicator of some hierarchy among the inhabitants, although this assumes that greater living space is an indicator of status. More sub-chambers could also reflect the presence of a larger number of attendants.
Finally, of course, as we have had occasion to note, the whole court could not have resided in the limited terrace area and within the harem itself space was at a premium. It is tempting to conclude that the mass ranks of the court generally resided in tents and covered wagons strewn about on the plain below and (those of highest rank perhaps) within the mud-brick and stone buildings on the plain, while the permanent stone buildings of the terrace, including the harem, were reserved for a privileged few of the inner royal family, with perhaps favoured wives and royal mothers the most likely to have their own apartments and, consequently, command the most intimate access to the king (Walthall 2008: 19).
World History
